Reducing the variability of the saponification step for MOSH&MOAH analysis using microwave-assisted extraction

The analysis of mineral oil hydrocarbons (MOSH&MOAH) in food poses challenges due to the chemical complexity of these contaminants, their similar chemical properties to the extraction matrix, and the high uncertainty associated with results, especially coming from different laboratories. This uncertainty is mainly associated with data integration and interpretation, accounting for roughly 20-30% of variability, which is summed to the variability due to sample preparation. The latter can easily exceed the 15% when fatty foods (as edible oils) are considered.

Recently, two interlaboratory trials focusing on fats and oils or on infant formula, reported an additional issue contributing to the uncertainty of the quantification, which is the discrepancy in the ratio between MOAH internal standards (specifically, 1,3,5-tri-tert-butylbenzene and methylnaphthalene). This discrepancy, attributed to varying compound partitioning during saponification/extraction, resulted in a ratio surpassing the expected 1.0. In both trials, the maximal tolerated ratio was 1.15, as reported in the updated Guidance for MOSH and MOAH analysis published by the Joint Research Center in 2023. However, this results in the acceptance of an additional 15% of variability on top of the ones mentioned above.

In this work, a different saponification method was optimized using a microwave-assisted extraction system to improve the reliability of the extraction step. Adjusting the solvent ratios used the ratio between the standards was reduced on average from 1.15 to 1.05 with good recoveries in different edible oils.